Proc. Nati. Acad. Sci. USA Vol. 88, pp. 5867-5871, July 1991 Biochemistry

Identification, cloning, and expression of a cytosolic megakaryocyte -tyrosine-phosphatase with to cytoskeletal protein 4.1 (cell signaling/protein phosphorylation) MINXIANG Gu, JOHN D. YORK, ILKA WARSHAWSKY, AND PHILIP W. MAJERUS* Washington University School of Medicine, Division of Hematology-Oncology, 660 South Euclid Avenue, Box 8125, St. Louis, MO 63110 Contributed by Philip W. Majerus, April 5, 1991

ABSTRACT We have isolated a cDNA encoding a third lular PTPases include PTPase 1B from human placenta (21, type of protein-tyrosine-phosphatase. We screened human 22), a related rat brain PTPase (23), and T-cell PTPase (24). megakaryoblastic cell line (MEG-01) and umbilical vein endo- The biological functions of soluble PTPases are not known, thelial cell cDNA libraries to obtain a 3.7-kilobase cDNA although overexpression of T-cell PTPase in BHK cells designated PTPase MEG. Northern blot analysis of MEG-01 reduces cellular protein phosphotyrosine (25). RNA detected a 3.7-kilobase transcript, suggesting that a We find that platelets contain very high levels of PTPase full-length cDNA has been identified. PTPase MEG cDNA activity. To characterize these enzymes we used a human contains an open reading frame of 926 amino acids. The cDNA megakaryoblastic cell line (MEG-01) cDNA as a template for has a G+C-rich 5' untranslated region of 771 nucleotides that polymerase chain reaction (PCR)-mediated amplification of has the potential to form stable stem-oop structures and has cDNA sequences between conserved heptapeptide se- two upstream ATG codons. The predicted protein (Mr = quences located in the catalytic domains of several PTPases. 105,910) has no apparent membrane-spanning region and In this way we identified several known PTPases and two contains a single protein-tyrosine-phosphatase domain (amino novel cDNA sequences (26). Here we report the sequence of acids 659-909) that is 35-40% identical to previously described a novel cDNAt that was used to screen MEG-01 cell and tyrosine-phosphatase domains. The recombinant phosphatase human umbilical vein endothelial cell (HUVEC) cDNA li- domain possesses protein-tyrosine-phosphatase activity when braries. The isolated full-length cDNA clone encodes a third expressed in Escherichia coli. The amino-terminal region (ami- type of PTPase, which contains a region that is similar to no acids 31-367) is 45% identical to the amino terminus of erythrocyte protein 4.1. human erythrocyte protein 4.1, a cytoskeletal protein. The identification of a protein-tyrosine-phosphatase that is related MATERIALS AND METHODS to cytoskeletal implies that cell signaling activities Human megakaryoblastic cell line MEG-01 was a gift from reside not only in transmembrane receptors but in cytoskeletal Hidehiko Saito (Nagoya University, Nagoya, Japan). The elements as well. HUVEC cDNA library was kindly provided by Evan Sadler (Washington University). Epidermal growth factor (EGF) Protein tyrosine phosphorylation is regulated by protein- was provided by Linda Pike (Washington Universi- tyrosine kinase (PTKase) and protein-tyrosine-phosphatase ty). The AZAP cDNA synthesis kit was purchased from (PTPase) activities. Since PTIKases are implicated in cell Stratagene. Sequenase, Taq polymerase, and 7-deaza-2'- growth and transformation (1, 2), PTPases must have related deoxyguanosine 5'-triphosphate were from United States Bio- functions (3-5). Nonproliferating cells such as blood platelets chemical. 32P-labeled deoxyribonucleotides were from ICN. and neurons contain high levels ofprotein tyrosine phosphate Deoxyadenosine 5'-[a-[35S]thio]triphosphate was from Amer- (6, 7), suggesting that tyrosine phosphorylation may serve sham. Oligo(dT)-cellulose was from Pharmacia. Restriction roles other than promoting cell growth. For example, throm- enzymes, phage T4 DNA ligase, and Klenow fragment of bin, collagen, and ionophore A23187 induce protein tyrosine DNA polymerase I were from Bethesda Research Laborato- phosphorylation in platelets (8-10); addition of vanadate and ries or New England Biolabs. Oligonucleotides were synthe- molybdate to electropermeabilized platelets increases tyro- sized on an Applied Biosystems 380B DNA synthesizer in the sine phosphorylation and stimulates secretion. These effects Washington University Protein Chemistry Facility. may result from inhibition of PTPases (11). PCR Amplification and Isolation of cDNA. A unidirectional Two types of PTPases have been defined: transmembrane MEG-01 cell cDNA library was prepared using a AZAP and intracellular. The transmembrane PTPases include leu- cDNA synthesis kit and 5 ,ug ofpoly(A)+ RNA from MEG-01 kocyte common antigen (LCA, CD45; ref. 12), LCA-related cells. The resultant library contains 3.6 x 106 independent molecule (LAR; ref. 13), Drosophila PTPase (DPTP; ref. 14), recombinants. Degenerate oligonucleotides A1, A2, and B several tyrosine phosphatases from human placenta (HPTP; were designed from sequences encoding amino acids con- ref. 15), three tyrosine phosphatases from human brain served among human PTPases (17). Sense oligonucleotides (RPTP; ref. 16), and LCA-related phosphatase (LRP; ref. 17). A1 and A2 were based on sequences WPDHGVP [5'- LAR may function in cell adhesion or in cell-cell interac- TGGCC(A or T)GA(C or T)CA(C or T)GGAGTCCCT-3'] tions, since its extracellular domain is structurally similar to and WPDFGVP [5'-TGGCC(A or T)GA(C or T)TT(C or the neural adhesion molecules N-CAM and fasciculin 11 (13). T-cell activation and proliferation require CD45 expression Abbreviations: PTKase, protein-tyrosine kinase; PTPase, protein- (18, 19). CD45 also activates a PTKase, pp56Ick, by dephos- tyrosine-phosphatase; LCA, leukocyte common antigen (CD45); phorylating a putative negative regulatory site (20). Intracel- LAR, LCA-related molecule; HUVEC, human umbilical vein endo- thelial cell; EGF, epidermal growth factor; RCM, reduced, carbox- amidomethylated, and maleylated. The publication costs of this article were defrayed in part by page charge *To whom reprint requests should be addressed. payment. This article must therefore be hereby marked "advertisement" tThe sequence reported in this paper has been deposited in the in accordance with 18 U.S.C. §1734 solely to indicate this fact. GenBank data base (accession no. M68941).

5867 Downloaded by guest on October 2, 2021 5868 Biochemistry: Gu et al. Proc. Natl. Acad. Sci. USA 88 (1991) A B Protein Tyrosine Phosphatase Domain (260a

FIG. 1. Diagram of a PTPase domain [260 amino LCA: WPDHGVPEDPHLLLKLRRRVNAFSNFFSG PIVVHCSAGVG acids (aa)] showing the positions of the oligonucleo- PTPase IB: WPDFGVPESPASFLNFLFKVRESGSLSPEHG PVVVHCSAGIG tides used to prime PCR. The novel sequences of two PTPase: WPDHGVPESPASFLNFLFKVRESGSLNPDHG PAVIHCSAGIG PCR products from the MEG-01 cDNA library and NOVEL-1: WPDHGVPDDSSDFLDFVCHVRNKRAGKEE PVVVHCSAGIG comparisons with the analogous regions of human NOVEL-2: WPDIFGVPSSAASL IDFLRVVRNQQSLAVSNMGARSKGQCPEPPIVVHCSAGIG PTPase 1B, T-cell PTPase, and LCA are shown.

T)GGAGTCCCT-3']. Antisense oligonucleotide B was based primers A1, A2, and B, we observed numerous PCR products on the sequence HCSAGIG [5'-(C or G)CC(A or G)AT- and 22 were subcloned and sequenced (Table 1). Sixteen of GCCTGCACT(A or G)CAGTG-3']. A template DNA from the clones were identified as previously known PTPases the MEG-01 cDNA library (50 ng) was amplified using Taq including T-cell PTPase, PTPase 1B, and LCA. The remain- polymerase in the presence of 300 ng of each oligonucleotide ing 6 encoded two sequences that we dubbed Novel-1 and as described (27). The PCR cycle, repeated 30 times, con- Novel-2 (Fig. 1). When we used the Novel-1 cDNA (probe I, sisted ofdenaturation at 940C for 1 min, annealing at 50'C for Fig. 2) to screen a MEG-01 cDNA library, the clone PTPase 1 min, and extension at 720C for 1 min. The PCR products MEG-1 was identified. This truncated clone encodes an were cloned into the Sma I site ofpBluescript SK(+) plasmid entire PTPase domain, a short 3' noncoding region, and a (Stratagene) (28) and sequenced. poly(A) tail (although no polyadenylylation signal is present). The MEG-01 cDNA library (2 x 106 recombinants) was screened with radiolabeled Novel-1 PCR product, and sub- P X HB K PEE N X sequently it and the HUVEC cDNA library (29) were S'~~~IIIII I~~~~~~~~~~~~~~ -(A)n 3' screened with various radiolabeled partial cDNA clones as Probe described (28). Nucleotide sequence was determined on both Novel-1 strands by the dideoxy chain-termination method. Northern Probe 11 blot analysis was carried out on MEG-01 RNA (28). PTFPase MEG-i1 Expression of Active PTPase in Escherichia coli. The 3' Probe III portion of PTPase MEG cDNA that encoded the PTPase PTPase MEG-14 E ~~~~~~~~~~~~~~.- 3 domain (bases 1264-2890) was excised from the PTPase MEG-14 clone. The resulting 1.6-kilobase (kb) cDNA frag- Probe IV ment was expressed in E. coli as described (30). Lysozyme PTPase HE-1 Probef was reduced, carboxamidomethylated, and maleylated Probe V (RCM lysozyme) according to Tonks et al. (21). RCM lyso- I PTPase HE-19 zyme (2 mg/ml) was then labeled on tyrosine by using [y-32P]ATP (1 mCi = 37 MBq) and EGF receptor kinase as described (21) except that the reaction mixture did not 0 1 2 3 3.7 Kb contain unlabeled ATP. PTPase activity was measured in 50 ,l of 10 mM imidazole/HCI (pH 7.2) containing 0.1% (wt/ = Band 4.1 homology domain M PTPase catalytic domain vol) bovine serum albumin and 0.1% (vol/vol) 2-mercapto- ethanol. After 10 min at 30°C, 17 Al of 2.5% bovine serum FIG. 2. The relationship of four cDNA isolates, a partial restric- albumin was added and protein was precipitated with tri- tion map of the composite full-length PTPase MEG cDNA, and the chloroacetic acid. Radioactivity in the supernatant was mea- probes designed for screening. The thin segments indicate noncoding sured with a liquid scintillation counter. sequences, and the thick segment indicates the open reading frame that encodes PTPase MEG. Selected restriction sites are shown: E, RESULTS EcoRI; B, BamHI; H, HindIII; K, Kpn I; N, Nde I; P, Pst I; X, Xho I. The thin open bars show the portion of the sequence contained in A PCR strategy was devised to identify different types of each ofthe cDNA isolates: PTPase MEG-1 and PTPase MEG-14 (from PTPases represented in the MEG-01 cell cDNA library. the MEG-01 library) and PTPase HE-i and PTPase HE-19 (from the Partially degenerate primers based on sequences that are HUVEC library). All clones were sequenced on both strands. highly conserved among PTPases but that flank a region of variable length and sequence were used to amplify cDNAs 1 2 for multiple enzymes in a single reaction (Fig. 1). Using Table 1. Number of independent cDNA subclones Apparent PCR No. of No. with identical DNA sequence product size, clones T-cell PTPase FIG. 3. Detection ofPTPase MEG base pairs sequenced PTPase IB LCA Novel-1 Novel-2 9.5- mRNA in MEG-01 cell RNA by blot 107 7 3 - 2 2 - 7.5 hybridization. Total (10 ug, lane 2) and lane 1) RNA 117 6 1 - 4 1 - poly(A)+ (2 ,ug, - rI samples from MEG-01 cells were -- - 4. 44 123 6 4 2 electrophoresed in a 2.2 M formalde- 153 3 - - 3 I.4 hyde/1% agarose gel and transferred 22 8 2 6 3 3 to a nitrocellulose filter. The cDNA 2.4-* insert of the PTPase MEG-1 clone PCR was carried out using primers A1, A2, and B with MEG-01 was used to probe the RNA blot. cDNA template. The products were size-fractionated by agarose gel Mobility of RNA standards (kb) is electrophoresis, eluted, subcloned, and sequenced. 1.44- shown at left. Downloaded by guest on October 2, 2021 Biochemistry: Gu et al. Proc. Natl. Acad. Sci. USA 88 (1991) 5869 This clone (probe II) was used to isolate PTPase MEG-14, a transcript detected by Northern blot hybridization of 2-kb clone. An 800-base-pair (bp) Xho I fragment from the 5' MEG-01 poly(A)+ RNA with probes II and III (Fig. 3). The end of this clone (probe III) was used to screen a HUVEC full 3661-bp nucleotide sequence and predicted amino acid library, yielding the 3-kb clone PTPase HE-1 (Fig. 2). When sequence of PTPase MEG are shown in Fig. 4. Base pairs we attempted to use probe IV to further screen the HUVEC -771 to 845 are from HUVEC cDNA clones while base pairs library, false positives that did not contain a PTPase domain 846 to 2890 are from MEG-01 clones. were obtained, so that a combination of probes III and V was The cDNA predicts a protein of 926 amino acids with Mr used for further screening to isolate PTPase HE-19. 105,910. Amino acids 31-367 of PTPase MEG ate 45% The structure of the composite 3.7-kb cDNA is outlined in identical to the amino-terminal region of human erythrocyte Fig. 2. Its length corresponds to the size of the mRNA cytoskeletal protein 4.1 (Fig. 5). The protein is also related to

,GCAGGGTTTTGGCCTCGCCTCTCGCGAGATCGCCTCCTGTTGCTGC -673

CGCCGCCGCTCCTGGCCACTGACTGGCGGCGCCTGCGCAGCCGCCATGTTCGG,TTGCTATGCTGCGGCCTAGGAGAGGGGGTGTGCTTGAGGGAGGAGG -574

rGACTGAGCTCCTCTTGCACTCTCACACACAAACGCTGCCCAGGATT -475

ACCCGCCAGCTCACGCCGCGCAGTGCGCTTTTCCGCTCCTCGCGCCCCACCAC 'CAACATTGTTCTCTCAGGACTCCTGGGTCCCAGGGGTCGGAATTGG -376

GCCTGAGCGGGAGAGGAAAGAGACTTGGCTTTGGCCGCGGGGTCGGAGGATTG ;GGGCCAGGCCCCCTCCCCCACGCACTTTTGGGGGTGTGGATTATCT -277

CATCCCTGCAGGGAGGTAGGAGAGGTCGCCGGCTGCCCGCCTCCCTGCCACCT('CCCCAGCGGCGCCGGCCCGCGGCTGCCCAGCAGCA3SCAGGTGGTGC -178

TGGCGGCTCCGGGTCGTGGCGCGACCGCTGCGGCGGCGGCTGCTCGGGGGGCG(,CTGAGGTAGCCCCCCGGAGCGGCACGGAGGACGCGCTTCTCCTCTG -79

CGCGCCGGGGCCTCdAGGCTTTTTTTCTCCAGCCGAGAGGACGCGGCTGTGATiATACGAAGACTTTGTGTGGACAGTAATGACCTCACGTTTCCGATTG 21 1 M T S R F R L CCTGCTGGCAGAACCTACAATGTACGAGCATCAGAGTTGGCCCGAGACAGACAGCATACTGAAGTGGTTTGCAACATCCTTCTTCTGGATAACACTGTA 120 8 P A G R T YN V R A S E L A R D R Q H T E V V C N I L L L D N T V CAAGCTTTCAAAGTCAATAAACATGATCAGGGGCAAGTCTTGTTGGATGTCGTCTTCAAGCATCTAGATTTGACTGAGCAGGACTATTTTGGTTTACAG 219 41 Q A F X V N K H D Q G Q V L L D V V F K H L D L T E Q D Y F G L Q TTGGCTGATGATTCCACAGATAACCCAAGGTGGCTGGATCCAAACAAACCAATAAGGAAGCAGCTAAAGAGAGGATCTCCTTACAGTTTGAACTTTAGA 318 74 L A D D S T D N P R W L D P N K P I R K Q L K R G S P Y S L N F R GTCAAATTTTTTGTAAGTGACCCCAACAAGTTACAAGAAGAATATACAAGGTACCAGTATTTTTTGCAAATTAAACAAGACATTCTTACTGGAAGATTA 417 107 V K F F V S D P N K L 0 EE Y T R Y Q Y F L Q I K Q D I L T G R L CCCTGTCCTTCTAATACTGCTGCCCTTTTAGCTTCATTTGCTGTTCAGTCTGAACTTGGAGACTACGATCAGTCAGAGAACTTGTCAGGCTACCTCTCA 516 140 P C P S N T A A L L A S F AV Q S E L G D Y D Q S E N L S GYL S GATTATTCTTTCATTCCTAATCAACCTCAAGATTTTGAAAAAGAAATTGCAAAATTACATCAGCAACACATAGGCTTATCTCCTGCAGAAGCAGAATTT 615 173 D Y S F I P N Q P 0 D F E K E I A K L H Q Q H I G L S P A E A E F AATTACCTAAACACAGCACGTACCTTAGAACTCTATGGAGTTGAATTCCACTATGCAAGGGATCAGAGTAACAATGAAATTATGATTGGAGTGATGTCA 7714 206 N YL N T A R T L E L Y G V E F H Y AR D 0 S N N E I M I G V M S GGAGGAATTCTGATTTATAAGAACAGGGTACGAATGAATACCTTTCCATGGTTGAAGATTGTAAAAATTTCTTTTAAGTGCAAACAGTTTTTTATTCAA 813 239 G G I L I Y K N R V R M N T F P W L K I V K I S F K CK Q FF I Q CTTAGAAAAGAATTGCATGAATCTAGAGAAACATTATTGGGATTTAATATGGTGAATTACAGAGCATGTAAAAATTTGTGGAAAGCATGTGTAGAACAT 912 273 L R K ELH E S R ET L L C F N M V N Y R A CK N L W K A C V E H CACACATTCTTCCGTTTGGACAGACCACTTCCACCTCAAAAGAATTTTTTTGCACATTATTTTACATTAGGTTCAAAATTCCGGTACTGTGGGAGAACT 1011 305 H T F F R L D R P L P P Q K N F F A HY F TL G S K F R Y C G R T GAAGTCCAATCAGTTCAGTATGGCAAAGAAAAGGCAAATAAAGACAGGGTATTTGCAAGATCCCCAAGTAAGCCCTTGCCACGGAAATTAATGGATTGG 1110 338 E V Q S V 0 Y GK E K A N K D R V F A R S P S K PL AR K L M D W GAAGTAGTAAGCAGAAATTCAATATCTGATGACAGGTTAGAAACACAAAGTCTTCCATCACGATCTCCACCGGGAACTCCTAATCATCGAAATTCTACA 1209 371 E V V S R N S I S D D R L E T Q S L P S R S P P G T P N H R N S T TTCACGCAGGAAGGAACCCGGTTACGACCATCTTCAGTTGGTCATTTGGTAGACCATATGGTTCATACTTCCCCAACCGAAGTGTTTGTAAATCAGAGA 1308 404 F T Q E G T R L R P S S V G H L V b H M V H T S P S E V F V N 0 R TCTCCGTCATCAACACAAGCTAATAGCATTGTTCTGGAATCATCACCATCACAAGAGACCCCTGGAGATGGGAAGCCTCCAGCTTTACCACCCAAACAG 1407 437 S P S S T 0 AN S I V L E SS P S Q E T P GD G K P P A L P P K 0 TCAAAGAAAAACAGTTGGAACCAAATTCATTATTCACATTCGCAACAAGATCTAGAAAGTCATATTAATGAAACATTTGATATTCCATCTTCTCCTGAA 1506 470 S K K N S W N Q I H Y S H S 0 0 D L E S H I N E T F D I PS S P E AAACCCACTCCTAATGGTGGTATTCCACATGATAATCTTGTCCTAATCAGAATGAAACCTGATGAAAATGGGAGGTTTGGATTCAATGTAAAGGGAGGA 1605 503 K P T P N G G I P H DN L V L I R M K P D E N G R F G FN V K G G TATGATCAGAAGATGCCTGTGATTGTGTCTCGAGTAGCACCAGGAACACCTGCTGACCTCTGTGTCCCTAGACTGAATGAAGGGGACCAAGTIGTACTG 1704 536 Y D Q K M P V I V S RV A P G T P A D L C V P R L N E G D Q V V L ATCAATGGTCGGGACATTGCAGAACACACTCATGATCAGGTTGTGCTGTTTATTAAAGCTAGTTGTGAGAGACATTCTGGGGAACTCATGCTTCTAGTT 1803 569 I N G R D I A E H T H D Q V V L F I K A S C E R H S G E L M L L V CGACCTAATGCTGTATATGATGTAGTGGAAGAAAAGCTAGAAAATGAGCCAGATTTCCAGTATATTCCTGAGAAAGCCCCACTAGATAGTGTGCATCAG 1902 602 R P N A V Y D VV E E K L E N E P D F 0 Y I P E K A P L D S V H Q GATGACCATTCCCTGCGGGAGTCAATGATCCAGCTAGCTGAGGGGCTTATCACTGGAACAGTCCTGACACAGTTTGATCAACTGTATCGGAAAAAACCT 2001 635 D D H S L R E S M I 0 L A E G L I T G T V L T Q F D Q L Y R K K P GGAATGACAATGTCCTGTGCCAAATTACCTCAGAATATTTCCAAAAATAGATACAGAGATA20CGCCTTATGATGCCACACGGGTCATTAAAAGGT2100 668 G M T M S CA K L P 0 N I S K N R Y R D I S P Y D A T R V I L K G AATGAAGACTACATCAATGCGAACTATATA2TATGGAAATTCCTTCTTCCAGCATTATAAATCAGTACATTGCTTGTCMGGGCCATTACCACACACT2199 701 N E D Y I N A N Y I N M E I P S SS I I N Q Y I A C 0 G P L P H T TGTACAGATTTTTGGCAGATGACTTGGGAACAAGGCTCCTCTATGGTTGTAATGTTGACCACACAAGTTGAACGTGGCAGAGTTAAATGTCACCAATAT 2298 734 C T D F W Q M T W E 0 G S S M V V M L T T Q V E R G R V K C H Q Y TGGCCAGAACCCACAGGCAGTTCATCTTATGGATGCTACCAAGTTACCTGCCACTCTGAAGAAGGAAACACTGCCTATATCTTCAGGAAGATGACCCTA 2397 767 W P E P T G S S S YGC Y Q V T C H S EE G N T A Y I F R K M T L TTTAACCAAGAGAAAAATGAAAGTCGTCCACTCACTCAGATCCAGTACATAGCCTGGCCTGACCATGGAGTCCCTGATGATTCGAGTGACTTTCTAGAT 2496 800 F N Q E K N E S R P L T Q I Q Y I A W P D H G V P DD S S DF L D TTTGTTTGTCATGTACGAAACAAGAGGGCTGGCAAGGAAGAACCCGTTGTTGTCCATTGCAGTGCTGGAATCGGAAGAACTGGGGTTCTTATTACTATG 2595 833 F V C H V RN K RA G K E E P V VV H C S A G I G R T G V L I T M GAAACAGCCATGTGTCTCATTGAATGCAATCAGCCAGTTTATCCACTAGATATTGTAAGAACAATGAGAGATCAGCGAGCCATGATGATCCAAACACCT 2694 866 E T A M C L I E C N Q P V Y P L D I V R T M R D Q RA M M I Q T P AGTCAATACAGATTTGTATGTGAAGCTATTTTGAAAGTTTATGAAGAAGGCTTTGTTAAACCCTTAACAACATCAACAAATAAATAAGAAAGCAAAAAG 2793 899 S Q Y R F V C E A I L K V Y E E G F V KP L T T S T N K * ATCTGGGATATGTGTTGGA2ACTGCTTTCCCTTATGTTCACTGTGCCATAATGCTGCTCGCAGGAAATGGCATTTTACAAAAAAAAAAAAAAAAA2890 FIG. 4. Nucleotide and deduced amino acid sequence of PTPase MEG. Amino acid sequence is shown in single-letter code below the nucleotide sequence. The major open reading frame begins at position 1. The first three in-frame ATG codons are boldface and the two short upstream open reading frames are underscored with broken lines. A sequence predicted to form a stem-loop structure is shown with two arrows. The sequence of the PCR product is underscored with a solid line. Numbers to the left refer to amino acid position, and those to the right indicate nucleotide position. Downloaded by guest on October 2, 2021 5870 Biochemistry: Gu et al. Proc. Natl. Acad. Sci. USA 88 (1991) ezrin (Fig. 5), a PTKase substrate (32, 33). Amino acids 659-909 of PTPase MEG are 35-40o identical to the phos- phatase domains of PTPase 1B, T-cell PTPase, LCA, and LAR. The region between these two domains (amino acids 368-657) does not match any sequences in the GenBank/ EMBL data base (January 1991). The 5' noncoding region of PTPase MEG cDNA was 771 bp and contains two short open reading frames predicting peptides of 8 and 22 amino acids (Fig. 4). The G+C content 1(1- of this region is 65% and the region -200 to -1 contains >80%o G+C. Several stable stem-loop structures have the potential to form in this region; most notable are 14-bp inverted repeats at positions -139 to -126 and -116 to -102 (Fig. 4). These can form stable stem-loops with a calculated Gibbs energy of AG37 = -26.3 kcal/mol (-0.7 kcal/mol per base) as analyzed using the FOLD program (34, 35). We searched GenBank and EMBL nucleic acid data bases with this sequence and found a 32-bp segment from the 5' un- translated region of the human protooncogene ABL (36) that Protei.n [Lt was 75% identical (data not shown). This region ofABL may FIG. 6. Dephosphorylation of [32P]phosphotyrosine-containing form a stem-loop structure similar to that in PTPase MEG, RCM lysozyme as a function of E. coli protein concentration. and 6 of 7 bases in the loop of the structures are identical. Various amounts of total homogenate from E. coli transformed with Expression ofPTPase MEG in E. coi. To determine that the pTrp or pTrp.PTPase MEG were incubated with 25,000 cpm of the protein encoded by PTPase MEG has PTPase activity, a RCM lysozyme substrate for 10 min. (Inset) Time course for de- 1.6-kb segment from the 3' end ofthe cDNA that encoded the phosphorylation of RCM lysozyme by the recombinant PTPase entire PTPase domain was subcloned into the bacterial MEG; 0.5 ,ug of total homogenate protein was incubated with 25,000 expression vector pTrp (30). A prominent 59-kDa protein was cpm of the RCM lysozyme substrate for various times. detected by SDS/PAGE oftotal, supernatant, and particulate fractions ofextracts ofE. coli transformed with pTrp.PTPase This protein, PTPase MEG, has a predicted Mr 106,000 and MEG (data not shown). About 90% of the recombinant can be divided into two functional domains, (i) a protein 4.1 protein was in the particulate fraction and was presumed to homology domain and (ii) a tyrosine-phosphatase domain, be in inclusion bodies. Extracts expressing pTrp.PTPase that are separated by 292 residues of unknown function. A MEG contained PTPase activity as measured with [32P]phos- hydropathy plot ofthe deduced amino acid sequence showed photyrosine-containing RCM lysozyme as substrate; this no hydrophobic regions that resemble a signal peptide or activity was inhibited by vanadate. Fig. 6 shows the time and transmembrane domain, and thus PTPase MEG is likely to be protein concentration dependence ofthe activity. Most ofthe a cytoplasmic protein. The tyrosine-phosphatase domain phosphatase activity was due to recombinant enzyme in the expressed in E. coli has activity, confirming that this cDNA supernatant fraction, since this fraction had twice the activity encodes a PTPase. The other intracellular PTPases that have of the particulate fraction (data not shown). been described all contain two functional domains: the ami- no-terminal phosphatase catalytic domain and the putative DISCUSSION carboxyl-terminal regulatory domain. The phosphatase do- We have isolated a cDNA from human MEG-01 cell line and mains of these proteins are 72-97% identical to each other HUVEC cDNA libraries that encodes a third type ofPTPase. (21-23). The phosphatase domain of PTPase MEG is only PTPase MEG 1MTSRFRLEAGRTY|NVRIEAS DEL RD ERO [HTEV [VNIL N 38 BAND 4.1 1 _ M HLmK V S LLDID 10 EZRIN . . K ...... 1 P NVRV T T M . . . . A 13 PTPasMEG 39 IT VIQANDJK- FND [ V Y F H LD T OID Y L L 74 BAND 4.1 11 TVI. .YE CVY E KHA KGCD R V c N L E EY F CL 46 EZRIN 144TE F3TOAT_P TT gW 0 V VMT I R E V H Y 48 PTPso MEG 75 A rf1D$TD NIEER L DIP N K PIRK Q L K R l8 P V L N F V KF 111 BAND 4.1 47WIDNA UTS K T . WLAK EW K O.[ J.[ IV W P 81 EZRIN 49 VK FT KLD KK SA E R REN PL 0KF AK FYP 85 MEG 112 V V F PTPaso S KFN K T R L QII 0D T RL P C P8 N T A A 147 BAND 4.1 I 82 PIDI -LEA OILITLD I T Ry Yl I~L OILSQRDV CsF ATA 117 EZRIN 86EEJV E EL I DL * LE0 E T 122 PTPss MEG 148 AYOSE L L DYDQE L Y ISF . . BAND 4.1 118 G D E D L IO180 QSE D P L F K .... EZRIN IL LS YIT L Il Y 150 123 L K C Y L E R VM D Q H 180 PTPasoMEG 181.. P Q D FrnKEmA K| HQ QR I GL S A EF LN T F T| 215 BAND 4.1 151 . . . T K E LJEIEKM ELH|K 8 YRLRJM T PA QAD L E N AK K L 185 EZRIN 161 KL T R D OWJDDP R VW AEAE R a MIL KD N M L E Y LIK I 0QDIL 198 PTPaseMEG 216 LIY1EF iH1Y[FAIRos S1N E I M 1 a I L I Y KIN R V[nRMNM T 252 BAND 4.1 186 XY G LIL H K E VV .LGVC $G ,4L VY D K L R I MR -222 EZRIN 199IM Y_qf N Y F E TFLgF K K T W L VID AL Lj tEK DD KL T P K 236 PTPaso MEG I F L 1LH 286 BAND 4.1 22325S3. FPWL VIK 8 CEFFIQ. RXf ErnRTL Ll F- LWKLI YKR8SF F I K.r GLjQ EW - ST I FI 256 EZRIN 2371 G PS E RI D P K A P D F . .. F 268 FIG. 5. Protein sequence re- PTPase MEG 287 NMV Y Y R A C NFWK KAC V E HFHT F F R D P . 0 N FF A 322 BAND 4.1 257 KL 8Sj RA WAK IVICVEHH T jFF RLITDTTD TJI LJ-. . . 290 lationship of the amino terminus EZRIN 269 Y APLR R N R I L 0 L C W EHR - -R K D T I E V 0 302 of PTPase MEG, human eryth- PTPass MEG 323 HYFrL C 8 K FERYIC VI 8 V Y CQKI ANN KD R AS PM 360 rocyte protein 4.1 (31) and hu- BAND 4.1 291 -FLaA K S R QT DRPAPE TA 327 man ezrin Identical EZRIN 3030 M K A A R ELHITI RIQIA8A HIF REK E (32). resi- - QL E R O L ETI KK R E TJE 337 dues are enclosed in boxes. PTPasa MEG 361 P L AIr 1 1 367 have BAND 4.1 328 KR A SIR ..I LI 334 Gaps been introduced to EZRIN 338 M M R E EELML R 349 optimize alignments. Downloaded by guest on October 2, 2021 Biochemistry: Gu et al. Proc. Natl. Acad. Sci. USA 88 (1991) 5871 38% identical to the other cellular PTPases in the catalytic 9. Golden, A. & Brugge, J. S. (1989) Proc. Natl. Acad. Sci. USA 86, domain. In addition, it has only 17 amino acids carboxyl to 901-905. 10. Nakamura, S. & Yamamura, H. (1989) J. Biol. Chem. 263, 7089- the phosphatase domain. Thus, the low degree of identity in 7091. the catalytic domain and the lack of a putative carboxyl- 11. Lerea, K. M., Tonks, N. K., Krebs, E. G., Fischer, E. H. & terminal regulatory domain indicate that PTPase MEG is only Glomset, J. A. (1989) Biochemistry 28, 9286-9292. distantly related to the other intracellular PTPases. In addi- 12. Ralph, S. J., Thomas, M. L., Morton, C. C. & Trowbridge, I. S. (1987) EMBO J. 6, 1252-1257. tion, the phosphatase domain of PTPase MEG is only 35- 13. Streuli, M., Hall, L. R., Saga, Y., Schlossman, S. F. & Saito, H. 40% identical to the phosphatase domains of the transmem- (1987) J. Exp. Med. 166, 1548-1566. brane-type PTPases. 14. Streuli, M., Krueger, N. X., Tsai, A. Y. M. & Saito, H. (1989) The amino-terminal similarity of PTPase MEG to protein Proc. Natd. Acad. Sci. USA 86, 8698-8702. 15. Krueger, N. X., Streuli, M. & Saito, H. (1990) EMBO J. 9, 4.1 suggests possible functions for the phosphatase. Protein 3241-3252. 4.1 has been identified not only in erythrocytes but also in 16. Kaplan, R., Morse, B., Huebner, K., Croce, C., Howk, R., Ravera, tissues such as brain (37), fibroblasts (38), blood platelets M., Ricca, G., Jaye, M. & Schlessinger, J. (1990) Proc. Natl. Acad. (39), and endothelial cells (40). It serves as a crosslinking Sci. USA 87, 7000-7004. 17. Matthews, R. J., Cahir, E. D. & Thomas, M. L. (1990) Proc. NatI. molecule with binding sites in the amino-terminal region that Acad. Sci. USA 87, 4444 4448. participate in the inositolphospholipid-dependent binding of 18. Pingel, J. T. & Thomas, M. L. (1989) Cell 58, 1055-1065. protein 4.1 to glycophorin, a transmembrane protein (41, 42). 19. Koretzky, G. A., Picus, J., Thomas, M. L. & Weiss, A. (1990) Since this is the region of similarity between PTPase MEG Nature (London) 346, 66-68. 20. Mustelin, T., Coggeshall, K. M. & Altman, A. (1989) Proc. Natd. and protein 4.1, PTPase MEG may also associate with a Acad. Sci. USA 86, 6302-6306. glycophorin-like transmembrane molecule. This suggests 21. Tonks, N. K., Diltz, C. D. & Fischer, E. H. (1988) J. Biol. Chem. that PTPase MEG has a cytoskeletal association with an 263, 6722-6730. additional PTPase activity that may function-in cytoskeletal 22. Chernoff, J., Schievella, A. R, Jost, C. A., Erikson, R. L. & Neel, rearrangement or organization. The central part of PTPase B. G. (1990) Proc. Natd. Acad. Sci. USA 87, 2735-2739. 23. Guan, K., Haun, R. S., Watson, S. J., Geahlen, R. L. & Dixon, MEG is unrelated to any previously sequenced protein and J. E. (1990) Proc. Natl. Acad. Sci. USA 87, 1501-1505. may interact with some other molecule to control the activity 24. Cool, D. E., Tonks, N. K., Charbonneau, H., Walsh, K. A., of the phosphatase domain. Fischer, E. H. & Krebs, E. G. (1989) Proc. Natd. Acad. Sci. USA PTPase MEG is related to other cytoskeletal elements, 86, 5257-5261. including ezrin (Fig. 5) and (28% identical over 191 25. Cool, D. E., Tonks, N. K., Charbonneau, H., Fischer, E. H. & Krebs, E. G. (1990) Proc. Natl. Acad. Sci. USA 87, 7280-7284. residues) (43). Ezrin was first identified as a major PTKase 26. Gu, M., York, J. D., Warshawsky, I. & Majerus, P. W. (1991) Clin. substrate in chicken microvilli and has been localized to Res. 39, 208a. microvilli in plasma membrane structures in a wide variety of 27. Saiki, R. K., Scharf, S., Faloona, F., Mullis, K. B., Horn, G. T., cell types (32, 33). After activation of A431 epithelial carci- Erlich, H. A. & Arnheim, N. (1985) Science 230, 1350-1354. noma cell EGF ezrin redistributes 28. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) Molecular receptors by EGF, rapidly Cloning: A Laboratory Manual (Cold Spring Harbor Lab., Cold into microvilli and membrane ruffles. The redistribution Spring Harbor, NY). correlates with cell surface shape changes and is coincident 29. Ye, R. D., Wun, T. C. & Sadler, J. E. (1987) J. Biol. Chem. 262, with an increased tyrosine and serine phosphorylation (44). 3718-3725. This suggests that ezrin is involved in cytoskeletal organiza- 30. York, J. D. & Majerus, P. W. (1990) Proc. Natl. Acad. Sci. USA 87, tion, and since ezrin is a PTKase substrate this organizational 9548-9552. 31. Conboy, J., Kan, Y. W., Shohet, S. B. & Mohandas, N. (1986) activity may be regulated by phosphorylation and dephos- Proc. Natd. Acad. Sci. USA 83, 9512-9516. phorylation oftyrosine residues. Many nonreceptor PTKases 32. Gould, K. L., Bretscher, A., Esch, F. S. & Hunter, T. (1989) are associated with membranes and some of them, including EMBO J. 8, 4133-4142. pp6v-src, pp120v--ab pp90V-gag-yeS pp80v-aYes, and type 33. Gould, K. L., Cooper, J. A., Bretscher, A. & Hunter, T. (1986) J. IV c-Abl, are bound to membrane cytoskeletons (45-47). Cell Biol. 102, 660-669. Several 34. Devereux, J., Haeberli, P. & Smithies, D. (1984) Nucleic Acids Res. cytoskeletal components have been identified as 12, 387-395. substrates for PTKases, including integrins, connexin, talin, 35. Freier, S. M., Kierzek, R., Jaeger, J. A., Sugimoto, N., Caruther, and vinculin (48). Since PTPase MEG is related to talin and M. H., Neilson, T. & Turuer, D. H. (1986) Proc. Natl. Acad. Sci. ezrin, we propose that its phosphatase activity is important USA 83, 9373-9377. in the regulation of cytoskeletal events. 36. Shtivelmen, E., Lifshitz, B., Gale, R. P. & Canaai, E. (1985) Nature (London) 315, 550-554. We thank Theo Ross and Anne Bennett Jefferson for useful 37. Goodman, S. R., Casoria, L. A., Coleman, B. D. & Zagon, I. S. discussions and Dr. Matt Thomas and Dr. J. Evan (1986) Science 224, 1433-1436. advice, Sadler for 38. Cohen, C. M., Foley, S. F. & Korsgren, C. (1982) Nature (London) critical review and comments, and Ann Delaney for assistance in the 299, 648-650. preparation ofthe manuscript. This research was supported by Grant 39. Davies, G. E. & Cohen, C. M. (1985) Blood 65, 52-59. HL14147 (Specialized Center for Research in Thrombosis), Grant 40. Leto, T. L., Pratt, B. M. & Madri, J. A. (1986) J. Cell. Physiol. 127, HL16634, and Training Grant HL07088 from the National Institutes 423-431. of Health. 41. Correas, I., Leto, T. L., Speicher, D. W. & Marchesi, V. T. (1986) J. Biol. Chem. 261, 3310-3315. 1. Yarden, Y. (1988) Annu. Rev. Biochem. 57, 443-478. 42. Correas, I., Speicher, D. W. & Marchesi, V. T. (1986) J. Biol. 2. Cantley, L. C., Auger, K. R., Carpenter, C., Duckworth, B., Chem. 261, 13362-13366. Graziani, A., Kapeller, R. & Soltoff, S. (1991) Cell 64, 281-302. 43. Rees, D. J. G., Ades, S. E., Singer, S. J. & Hynes, R. 0. (1990) 3. Gould, K. L. & Nurse, P. (1989) Nature (London) 342, 39-45. Nature (London) 347, 685-689. 4. Dunphy, W. G. & Newport, J. W. (1989) Cell 58, 181-191. 44. Bretscher, A. (1989) J. Cell Biol. 108, 921-930. 5. Hunter, T. (1989) Cell 58, 1013-1016. 45. Burridge, K., Fath, K., Kelly, T., Nuckolls, G. & Turner, C. (1988) 6. Brugge, J., Cotton, P., Lustig, A., Yonemoto, W., Lipsich, L., Annu. Rev. Cell Biol. 4, 487-525. Coussens, P., Barrett, J. N., Nonner, D. & Keane, R. W. (1987) 46. Drubin, D. G., Mulholland, J., Zhu, Z. & Botstein, D. (1990) Nature Dev. 1, 287-2%. (London) 343, 288-290. 7. Golden, A., Nemeth, S. P. & Brugge, J. S. (1986) Proc. NatI. Acad. 47. Van Etten, R. A., Jackson, P. & Baltimore, D. (1989) Cell 58, Sci. USA 83, 852-856. 669-678. 8. Ferrell, J. E., Jr., & Martin, S. (1988) Mol. Cell. Biol. 8, 3603-3610. 48. Bishop, J. M. (1991) Cell 64, 235-248. Downloaded by guest on October 2, 2021